Breaking the accuracy-resource dilemma: a lightweight adaptive video inference enhancement

Wei Ma, Shaowu Chen, Junjie Ye +2

Existing video inference (VI) enhancement methods typically aim to improve performance by scaling up model sizes and employing sophisticated network architectures. While these appr…